What is Sume? A video agent platform, its API, and billing
Sume is a video agent platform: brief an agent in chat, save the recipe as a Format, and call it from your backend over one API. Surfaces and billing.

Sume is a video agent platform. People author and iterate on video with an agent in chat, save the result as a reusable recipe called a Format, and call that Format from their own backend over HTTP. Individual generation tools, such as image, video, avatar, text-to-speech, and timeline, also exist as HTTP APIs.
That definition is from Sume basics, read on 2026-09-27; the public site titles the product "Video Agent API built for automation." This page lists only what the docs and site say Sume offers. What is a video agent? explains the idea behind it.
What does Sume offer?
The product map in Sume basics lists these surfaces:
- Agents: the chat UI at
www.sume.com/agents, where a person writes a brief, approves spend, inspects artifacts, and saves a recipe. - Formats: saved recipes your backend calls by handle and slug. Each run boots a fresh sandbox, loads the recipe, runs the agent with generation tools, and returns artifacts plus optional structured JSON. Ready-made Formats by Sume answer at the
sumehandle. - Agent Completions and Scheduled: the same agent on an ad-hoc task from your backend, or a saved task on a cadence.
- Models: single generation endpoints, such as
POST /v1/videos(a catalog model orsume/auto),POST /v1/images, Sume Avatar 1.0, and the Music Router. - A TypeScript SDK,
@sume-com/sdk, and a dashboard for keys, jobs, usage, and billing.
What are Sume's domains and entry points?
The docs call the Format API the surface most partners should integrate; models are supporting building blocks for when you need one invocation and nothing else. A CLI and hosted MCP also exist and still work, but the docs say they are not the primary path today.
| Entry point | Address | What it is |
|---|---|---|
| Developer API | https://api.sume.com/v1 | Public, workspace-scoped, key-authenticated API |
| OpenAPI | https://api.sume.com/reference/json | Live schema, the source of truth for fields |
| Media | https://media.sume.com | First-party generated media artifacts |
| Agents | https://www.sume.com/agents | Chat authoring UI |
| Dashboard | https://www.sume.com/dashboard | API keys, jobs, usage, billing & subscription |
| Hosted MCP | https://mcp.sume.com/mcp | Remote MCP endpoint; not the primary path today |
| Docs | https://docs.sume.com | Guides and the API reference |
How do developers call Sume?
Create an API key in the dashboard and send it from a server-side environment as Authorization: Bearer or x-api-key. The key carries the workspace, so you do not send a workspace id in request bodies. A Format run is one POST /v1/formats/{handle}/{slug}/runs that answers with a receipt; you then take a signed webhook or poll. Sume API quickstart walks through a first call, and What is a Sume Format? covers the recipe.
How does Sume charge?
Plans are Free $0, Pro $40 / month, Startup $120 / month, Scale $400 / month, and Enterprise covers custom terms or invoice billing. Usage is billed at each model's published rate plus a 5.5% agent fee by default, and video generation, the Sume Agent, Formats, and the API all draw from one balance.
- API pricing is the public metered rate card and the source of truth for prices.
- You buy credits on the dashboard's Billing & subscription page; the public API reads
GET /v1/balanceandGET /v1/usagebut has no top-up endpoint. - Every Format run carries a generation spend cap, and each receipt reports the cap and what the run spent against it. How Sume pricing works and What is the Sume agent fee? go further.
Where do outputs and keys live?
API keys and spend resolve to a workspace, personal or team. A team-owned Format needs a key created in that team's workspace. Generated outputs that belong to Sume come back as media.sume.com URLs, which the docs call the public artifact contract; Do AI-generated video URLs expire? covers how long they last and who can open them.
Sources
Related posts
More in Agents
- YouTube chapter generator: timestamps from a transcript
A YouTube chapter generator turns a transcript into timestamps and titles from 00:00: sentence timings from speech-to-text, topic breaks from an LLM.
- Agentic video editing: what an AI editing agent does
Agentic video editing means an AI agent plans an edit, runs it with editing tools, and checks the result. What those tools are and what they can read.
- AI agent vs workflow: the difference and when to use each
A workflow runs steps fixed in advance; an AI agent decides its own steps. When each fits, what an agentic workflow is, and how Sume offers both.
- Are MCP servers only for Claude? No, other apps use them too
No. MCP is an open standard, so ChatGPT, Codex, Cursor, VS Code, and Gemini CLI can use the same server. What differs is each client's config and sign-in.
Written by Sume